GPS is a really cool invention but what if you could position yourself in a building using the dozens of WiFi access points positioned throughout the campus!
What it does
First it records access point signal strength data from various points throughout the building and stores them in an SQLite database within the Android app. When it's ready to determine a location, it again takes the signal strength of all nearby access points and compares it to the measurements of the points recorded.
How we built it
Our app consists of two pieces, the Wifi Recorder which records all the WiFi access points located by the device. It stores signal strength and a text value representing the location in an SQLite database. This was exported to a CSV file.
We built another application that would take the CSV file and calculate the averages and standard deviations of the tens of thousands of data points for each MAC Address.
The second part is the Indoor GPS which imports the data generated from the CSV and then looks at the points surrounding it to guess where the user is now.
Challenges we ran into
We had trouble using content providers and a few other elements of the Android API. Although all the logic works, we've consistently ran into minor bugs that prevented us from having a full locator prototype in 24 hours.
What's next for ITE Positioning System
We hope to polish the technology so it works throughout all of ITE, rather than just a few select rooms. We also hope to in the future adopt more complex machine learning techniques to find a users location based on collected data. We also hope to simplify the two apps into a single app.